Active contour for Intelligent Road Tracking System

Abstract

In this paper a computationally low cost, real time vision based tracking system is proposed for autonomous guided vehicle which is currently under development at the URRG Lab. In this paper, two major aspects for our tracking system are discussed. First aspect is the simple road tracking algorithm using the active contour. A quadratic B-spline function is used to model the road boundary then image processing is applied along the curve normal for feature point extraction followed by the recursive curve fitting technique for pose and orientation estimation. The second aspect is the development of a low cost and flexible hardware for the visual tracking system and its real time analysis.